The long-term goal of this proposal is to increase intrahepatic and lung tumor dose and intrahepatic tumor resectability by optimizing dose distributions within the tumor and individualizing assessment of normal tissue toxicity though innovative normal tissue imaging. This goal will be carried out through two specific aims. Aim 1 is to optimize the use of radiation therapy in the treatment of intrahepatic cancer. We will escalate high dose non-uniform radiation with the goal of increasing both local control and resectability. During treatment we will measure global changes in hepatic function using indocyanine green extraction, and regional changes in blood flow using dynamic contrast-enhanced CT scanning to assess early liver injury. These data will permit us to develop a model in which early changes in liver function can be correlated with liver toxicity after treatment. We also propose to use this model to develop a new protocol in which we reoptimize planning during the course of treatment to minimize normal liver injury. Aim 2 is to optimize the use of radiation therapy in the treatment of Stage III unresectable lung cancer. We will assess the optimal timing of chemotherapy and escalated radiation in a randomized phase II trial. During treatment we will measure changes in FDG-PET, to assess early tumor response, and full ventilation-perfusion SPECT scanning and pulmonary function tests to assess early lung injury. These data will permit us to develop a model in which early changes in FDG-PET imaging and lung function can be correlated with tumor response and lung toxicity, respectively, after treatment. We also propose to use this model to develop a new protocol in which we re-optimize planning during the course of treatment to maximize tumor control and to minimize normal tissue injury. We anticipate that these studies will increase the local control and resectability of intrahepatic cancers and the local control of lung tumors while individualizing and, thus, minimizing the risk of normal tissue injury.